Introdução: Apesar de ser uma intervenção potencialmente promissora para melhorar a cognição em sujeitos com comprometimento cognitivo leve (CCL), os efeitos do treinamento aeróbio (TA) no metabolismo regional de glicose (MRG) no cérebro ainda são pouco conhecidos. Objetivo: Avaliar os efeitos do TA na cognição e no MRG cerebral em sujeitos com CCL, usando tomografia por emissão de pósitrons com 18F - fluordesoxiglicose (PET-18FDG). Métodos: Sessenta e cinco sujeitos com CCL realizaram um programa de TA de intensidade moderada, duas vezes por semana, durante 24 semanas. Uma avaliação cognitiva com a escala Alzheimer\'s Disease Assessment Scale - Cognitive Subscale (ADAS-Cog) e outra do metabolismo cerebral em repouso com o PET-18FDG foram realizadas antes e depois da intervenção. A capacidade aeróbia foi comparada utilizando o consumo máximo de oxigênio (VO2 máx.) medido em mL/kg/min. Os dados de PET-18FDG foram analisados voxel por voxel com o programa SPM8 (limiar: p < 0,001 não corrigido, tamanho de agrupamento >= 50 voxels). Uma técnica de modelação de equações estruturais foi usada para criar um modelo para explorar as mudanças no MRG cerebral em repouso depois do programa de TA. Resultados: Comparações com testes t pareados revelaram melhora nos valores da escala ADAS-Cog (diferença média: -2,4 (3,5), p < 0,001) e no VO2 máx. (diferença média: 1,4 (2,3) mL/kg/min, p < 0,001). A análise metabólica do cérebro constatou uma redução bilateral do MRG em ambos os polos temporais anteriores, no giro frontal inferior esquerdo, no giro do cíngulo anterior esquerdo, no hipocampo direito, no giro frontal médio esquerdo e nos núcleos caudados bilaterais. Em contraste, houve um aumento do MRG no precuneus direito e no giro frontal inferior esquerdo. Um modelo de confirmação de fatores foi usado para criar variáveis latentes que refletem a rede de saliência (RS) e a rede de modo padrão (default mode network (DMN)) com base em regiões com mudanças no MRG encontradas depois do TA. Uma análise de caminhos foi utilizada para testar a hipótese de que o TA poderia induzir mudanças (diminuição) do MRG em regiões da RS. Essa redução pode refletir um aumento da eficiência da rede e, consequentemente, um melhor controle sobre outras redes. Em última instância, essa mudança induziria um aumento do MRG no precuneus, um importante nó da DMN. Os valores de adequação do modelo atingiram significância estatística, demonstrando que esse modelo explica a variância dos dados medidos. Conclusão: O TA melhorou a cognição e alterou o MRG em regiões da DMN e da rede de saliência em sujeitos com CCL / Background: Despite being a potentially promising intervention to improve cognition in subjects with mild cognitive impairment (MCI), the effects of aerobic training (AT) on regional brain glucose metabolism (rBGM) are not yet entirely understood. Objective: To evaluate the effects of AT on cognition and rBGM in subjects with MCI using 18FDG-PET. Methods: Sixty-five subjects with MCI performed a twice-a-week, moderate intensity, AT program for 24 weeks. Assessment with the ADAS-Cog (Alzheimer\'s Disease Assessment Scale - Cognitive Subscale) and evaluation of rBGM with 18FDG-PET were performed before and after the intervention. Aerobic capacity was compared using the relative maximal oxygen consumption (VO2 max) measured in mL/Kg/min. 18FDG-PET data were analyzed on a voxel-by-voxel basis with SPM8 (threshold: uncorrected p < 0.001, cluster size >= 50 voxels). Structural equation modeling was used to create a model to explore the changes of resting-state rBGM after the AT program. Results: Comparisons using paired t-tests revealed improvements in the ADAS-Cog (mean difference: -2.4 (3.5), p < 0.001) and VO2 max scores (mean difference: 1.4 (2.3) mL/kg/min, p < 0.001). Brain metabolic analysis revealed a bilateral reduction of rBGM in both anterior temporal poles, left inferior frontal gyrus, left anterior cingulate cortex, right hippocampus, left middle frontal gyrus and bilateral caudate nuclei. In contrast, there was an increase in rBGM in the right precuneus and left inferior frontal gyrus. Confirmatory factor analysis was used to create latent variables reflecting the salience network (SN) and default mode network (DMN) based on regions with rBGM changes found after the AT. A path model analysis was used to test the hypothesis that AT would induce rBGM changes (decrease) in regions of the SN. This decrement may reflect improved efficacy and consequently, better control over other networks. Ultimately, these changes would induce an increment in rBGM in the precuneus, an important node of the DMN. Values of model fit reached statistical significance, demonstrating that this model explain the variance of the measured data. Conclusion: AT improved cognition and changed rBGM measured with 18FDG-PET in nodes od the DMN and SN in subjects with MCI

En raison de sa localisation en profondeur du cerveau, le rôle de l’insula dans le traitement de l’information est longtemps resté énigmatique. Or, l’avènement des techniques de stimulation électro-corticale et de neuroimagerie a permis de mettre en exergue son implication dans divers aspects du fonctionnement neuropsychologique. De plus en plus d’études suggèrent que le cortex insulaire joue un rôle clé dans le traitement des caractéristiques physiques des stimuli sensoriels, ainsi que dans le traitement de la saillance des informations. Les théories contemporaines avancent ainsi que l’insula serait une région cruciale dans le « réseau de saillance » et serait impliquée dans les processus sensoriels, émotionnels et attentionnels. Toutefois, la nature exacte de sa contribution demeure inconnue, notamment en raison des limitations intrinsèques des techniques d’investigation traditionnelles, ainsi que de la faible prévalence des lésions circonscrites à l’insula, d’autant que l’évidence clinique ne fait pas l’unanimité. En outre, les résections insulaires sont de plus en plus fréquentes chez les patients atteints d’épilepsie insulaire pharmaco-résistante. Cependant, les altérations neuropsychologiques d’une telle intervention restent mal connues. Ainsi, les études qui composent cette thèse visent à mieux comprendre la façon dont l’insula participe au traitement de l’information et les conséquences neuropsychologiques des résections insulaires sur les processus sensoriels, émotionnels et attentionnels. Les deux premières études de cette thèse documentent les contributions respectives des portions antérieure et postérieure de l’insula au traitement attentionnel pour l’information sensorielle. Les réponses de l’insula lors de l’exécution de tâches attentionnelles de type oddball visuel et auditif sont enregistrées au moyen de l’EEG intracrânienne auprès de patients atteints d’épilepsie dont des électrodes ont été implantées dans l’insula dans le cadre d’une évaluation préchirurgicale pour une épilepsie résistante à la médication. Les résultats suggèrent que l’insula antérieure participe au déploiement attentionnel volontaire aux alentours de 300-500 ms à la suite de la présentation de stimuli pertinents à la tâche en modalité visuelle et auditive, alors que la portion postérieure, quant à elle, est impliquée dans le traitement attentionnel automatique survenant de manière précoce, autour de 100 ms suivant la présentation d’informations auditives, indépendamment de la pertinence du stimulus. Les deux études suivantes qui composent cette thèse examinent les conséquences neuropsychologiques d’une résection au cortex insulaire sur le traitement sensoriel et affectivo-attentionnel, chez des patients épileptiques réfractaires à la médication qui ont subi une résection unilatérale de cette région. Leurs performances dans une tâche Dot-Probe révisée et dans un test de Stroop émotionnel, ainsi que leurs réponses à un questionnaire mesurant des patterns comportementaux sensoriels, sont comparées à celles d’un groupe de patients ayant subi une chirurgie d’épilepsie temporale et d’un groupe d’individus contrôles en santé. Les résultats mettent en évidence des altérations sensorielles et du contrôle des interférences émotionnelles à la suite d’une chirurgie d’épilepsie insulaire. En somme, les données de cette thèse contribuent à une meilleure compréhension du rôle spécifique de l’insula au traitement de l’information sensorielle, saillante, émotionnelle et attentionnelle, au moyen de mesures neurophysiologiques et comportementales. Elles fournissent également un appui quant à la pertinence de développer des outils standardisés en évaluation neuropsychologique afin de mieux identifier les perturbations fonctionnelles associées à une épilepsie ou une chirurgie d’épilepsie insulaire. / The role of the insular cortex in information processing has long been considered enigmatic, partly due to its deep location in the brain. However, the advent of direct electrocortical stimulation and neuroimaging approaches have shed light on its involvement in multiple of neuropsychological functions. An increasing number of studies suggest that the insular cortex plays a crucial role in processing the physical characteristics of sensory stimuli, as well as in the processing of salient information. Current theories argue that the insula would be a critical structure in the “salience network” and involved in sensory, emotional and attentional processes. However, the specific contribution of the insular cortex remains unknown, notably due to the intrinsic limitations of conventional approaches and the very low prevalence of lesions restricted to the insula, especially as little clinical evidence support these findings. Furthermore, although insular resections are becoming more frequent, the neuropsychological effects of this surgery remain unclear. Thus, the studies that make up this thesis aim to improve our understanding of the role played by the insula in the salient information processing and the neuropsychological consequences of the insular resections on sensory, emotional and attentional functions. The first two studies of this thesis assess the respective contributions of the anterior and posterior insular portions in attentional processing towards salient and relevant sensory information. The insular responses during visual and auditory oddball attentional tasks are recorded by means of intracranial EEG (iEEG) in epileptic patients undergoing invasive iEEG, with electrode contacts implanted in the insula as part of a pre-surgical evaluation of their drug-resistant seizures. The results suggest that the anterior insula participates in voluntary attentional processing around 300-500 ms following the presentation of task-relevant stimuli in both visual and auditory modality, whereas the posterior portion is involved in automatic processing occurring about 100 ms after auditory stimuli presentation, independent of task-relevant information. The next two studies examine the neuropsychological consequences of insular cortex resections on the sensory and affectivo-attentional processes, in drug-refractory epileptic patients who have undergone unilateral resection of this structure. Their performance in a revised Dot-Probe task and an emotional Stroop test, as well as their responses in a questionnaire assessing sensory behavioral patterns, were compared to a group of patients who had surgery for temporal lobe epilepsy and a group of healthy control. The results highlight alterations in sensory processing and emotional interference control following insular epilepsy surgery. In sum, the neurophysiological and behavioral data in this thesis contribute to a better understanding of the specific role of the insula in the processing of sensory, salient, emotional and attentional information. Moreover, these findings highlight the need to further develop neuropsychological tests in order to better identify functional disturbances associated with insular epilepsy and insular resection surgery.